Tree calculus is minimal, modular, Turing-complete, reflective anyone has any opinions on this?

I spent a little time with this but didn't really manage to understand anything of it. I buy that it's Turing complete, but it's unclear what its benefits are.

It's interesting and elegant, but without an algorithm to turn a tree calculus program into a performant executable, it's limited in usefulness for a GP programming language or compiler target/source. It does seem like it could be useful for analyzing programs though, depending on what properties of the programs you're looking to explore/analyze. Sadly, the lack of a normal form for all programs is a little limiting; if one existed this language would be EXTREMELY useful (example: could be used to prove any compiler optimization produces equivalent programs). I do wonder if there is some other similar calculus that's a little less powerful but more useful-- This approach of developing a calculus to greatly simplify a complex world has worked great in robotics (see https://bivector.net/doc.html for fun applications of geometric algebra)